Re: Profiling and certificates for MathML. Avoiding imitators

William F Hammond wrote:

> Yes, Mozilla discourages redundant mrows; but, no, they do not cause
> real harm,

Could you guarantize this on official paper?

> and they are sometimes a necessary side effect of automatic
> translation from an author-level markup to XHTML+MathML.

Necessary side effect? I can understand from {where} those <mrow> arise in
a TeX like input, but cleaning any extra <mrow> containing a single child
is not a hard problem.

>> Invisible times continues being represented by justaposition.
>
> This is a controversial matter; please do not pretend otherwise.

What is controversial? The invisible times concept? The MathML entity? The
Unicode character?

> In mathematics juxtaposition usually has a default meaning as an
> operation that is context-dependent.

Visual juxtaposition? I was refering to juxtaposition at a code (markup)
level.

I think that 'all' [1] computer languages and content/semantic markup or
knowledge representation system do not use justaposition for infering
meaning multiplication. In Python, one writes [4 * pi * j]; not [4 pi j].

So far as I know no computer is intelligent enough to differentiate [L
rho] from [L rho] speaking correctly each one, can you?

[4 pi i] as encoded in that blog can be easily reproduced even with HTML 4
code. It looks as using quantum mechanics for computing trajectories for a
Boeing.

> With presentation markup it is a matter for the author (and
> the editor) to settle so long as the document is valid.

Agree, no less true that <span>settle to long so the as is document
valid</span> is correct HTML too. And nobody want a web full of that.

> If an author does not want to use &invisibleTimes, the effect
> of labeling his/her markup wrong is to drive him/her away from
> placing XHTML+MathML content on the web.  Is that what we want?

Good question. I got no problems with a personal page being written by a
12 years-old schoolboy, but I referencied an ultratechonological blog
(self-claiming to be the most advanced of the world) devoted to academic
stuff (string theory and others) and being driven by a physicist. either
you asumme that academic will be stuff correctly or you need some kind of
formal certification.

We have peer-review on formal publishing at journals, therefore first
option appears to be not enough.

We already defined certifications thecniques for languages of markup as
CML (that Gkoutos, Murray-Rust, Rzepa, and Wright call the "Signed
Chemical Web Of Trust").

Even the Wiki has some kind of minimal recommendations for input content,
and editors/managers periodically check articles on the enciclopedia,
eliminating low-level stuff. Why would MathML be different, specially at
academic level of communication?


[1] Mathematica, I think, leaves that as option providing explicit
multiplication.

Received on Thursday, 21 December 2006 18:33:46 UTC